Thomas Philippon (NYU)
Thomas Philippon is the Max L. Heine Professor of Finance at New York University, Stern School of Business. Professor Philippon was named one of the “top 25 economists under 45” by the IMF in 2014. He has won the 2013 Bernácer Prize for Best European Economist under 40, the 2010 Michael Brennan & BlackRock Award, the 2009 Prize for Best Young French Economist, and the 2008 Brattle Prize for the best paper in Corporate Finance. Professor Philippon has studied various topics in macroeconomics and finance: systemic risk, crisis resolution mechanisms, the dynamics of corporate investment and household debt, and the size of the finance industry. His recent work has focused on the Eurozone crisis, financial regulation, and the market power of large firms.
On 3rd August 2021, Professor Thomas Philippon joined us in a Luohan Academy Webinar to discuss the dynamics behind the data sharing between merchants and a platform, and its consequential impacts on consumer welfare and market power. Professor Philippon shared his recent co-authored paper "Data Sharing and Market Power with Two-Sided Platforms", which tries to answer the following three questions:
(i) How does the platform's data gathering affect buyers and sellers?
(ii) Is there a role for regulating data collected by platforms?
(iii) How do new internet platforms differ from the traditional ones?
The paper begins with a model for the traditional market where a platform is absent. In the traditional market model, the supply side consists of legacy merchants and new merchants selling various goods, and the demand side consists of consumers with a specifically assumed preference for the goods. There are two decisions to make in the model. One is for the new merchants, they need to decide whether to pay the entry cost and enter the market in stage one. After their decisions, the supply side is clear, and the other decision goes to the demand side. Consumers are matched to sellers in this market, and they need to decide whether to purchase goods based on their preferences in stage two. The equilibrium of this model is fully characterized by a notation called market tightness, which is indeed a seller's probability of being matched to a consumer. The take-away of this model is that the equilibrium for the traditional market is efficient as the solution to a social planner's decision problem coincides with the solution to merchants' profit maximization problems.
Starting from this model of the traditional market, the paper proceeds to add the two-sided platform to the market. The platform provides two technologies: a matching technology and an information technology. The matching technology is the same as the one that a traditional market has, while the information technology provides merchants a signal of consumer's preference. Merchants can learn consumers' preferences in a Bayesian way and change the variety of goods they sell. The timing of this model with a platform is as follows. In stage one, consumers decide to participate on either the platform or the outside market, and if they participate onthe platform they choose a disclosure policy that sends their preference to the platform as a signal. In stage two, merchants make the same decision of entry as in the traditional model. As for stage three, merchants and the platform bargain on trading consumer's data, and if no deal closes, the merchant can sell in the outside market. The final stage has two venues, merchants who stay on the platform receive consumer's data and are matched to consumers. Merchants who sell in the outside market are matched to consumers without consumer's data.
The paper then uses these two models to answer the questions asked in the introduction. It finds that data sharing increases gains from trade by improving match quality, but also increases the market power of platforms relative to sellers. There are two externalities which increase the market power of the platform and lower merchant entry. The gatekeeper externality arises from the platform's control over access to consumers. The copycat externality derives from its ability to compete with its own merchants. These effects are not internalized by consumers and data sharing can be socially excessive. Hence if there is too much data, the role we need is to use regulation to push down these two externalities so as to reduce platform market power relative to sellers. On the other hand, if there is too little data, the role we need is to facilitate data collection so as to incentivize seller entry which benefits consumers. In the end, the difference between new internet platforms and traditional ones is mainly the efficiency of the matching function and data gathering.
Throughout Professor Philippon's presentation, other participants including Liyan Yang from the University of Toronto, Kose John from the New York University, and Luohan Academy economist Xijie Gao discussed different parts of this paper. For example, Xijie Gao asked that from a social point of view, do we really want the market power to be shared between sellers and platforms? Is it possible that market power on the platform side could be better? This question triggered Professor Philippon to share one missing part of this paper: platform entry. Professor Philippon mentioned market power on the platform side is bad if we only have one platform, which is the assumption of this paper. However, this conclusion might not hold if we have a market with multiple competing platforms and the new conclusion depends on the matching and information technology each platform has. The Q&A session was concluded by comments from Long Chen.
If you would like to give a presentation in a future webinar, please contact our Economist Dr. Xijie Gao (email@example.com).